In some domains, the call to break this vicious cycle is already happening. E.g. ACM Sigcomm is one of the most prestigious and exclusive conference in Networking and Distributed Systems research. Some of the most profound researchers in that domain are driving a pledge to fundamentally rethink what the conference accepts and what gets presented: https://sigcomm.quest/proposal.html.
This is a terrible idea. People will have friends who get their papers in, in return for the favor of the same.
Agree with your general point though. No easy answers though.
But this is already the case. In some fields even double blind doesn't work because community members are aware of each others research. If you're one of two people worldwide who's specific specialization is the compression of semantic knowledge streams, you're probably willing to be money on your ability to spot your colleagues research.